The Revelation of Insider Trading using the Spread: A Comparison with the Sports Betting Market Christophe Barraud1 and Philippe Gillet2 Abstract This article compares methodologies which can be used to detect insider trading on financial markets and frauds on sports betting industry. There is a lot of evidence coming from the literature showing the similarities on financial markets and sports betting market. Furthermore, some authors highlight a spread enlargement depending on the proportion of uninformed investors during insider trading on the financial markets. Considering that, we used the bid-ask spread to see if such a phenomenon exists on sport betting industry. We used a sample of 10 litigious tennis games compared with a control-sample to examine spread’s behavior in the betting industry. We find out that, as on financial markets, spreads of litigious games increase on gambling markets before the beginning of the match, when the proportion of uninformed bettors is low. So, we can make two conclusions: on the one hand, the spread behavior could be used to detect frauds on gambling markets. On the other hand, we confirm spread’s behavior during insider trading on financial markets, which is comforting the strong hypothesis of the market efficiency theory. Keywords: Market efficiency, betting industry, spread, insider trading, strong hypothesis. Introduction The recent opening of online sports betting in Southern Europe and particularly in France, coinciding with 2010 Soccer World Cup, gives a new visibility to this activity. It becomes difficult in these countries to consult a website without being solicited by advertisements for 1 University of Paris-Dauphine, Paris, France 2 University of Paris-Sud, Paris, France Corresponding author : Philippe Gillet, Email : [email protected] 1 online gambling sites. However, this activity has long existed in English speaking countries, including Britain, where the art of bookmaking is integral part of British culture. In fact, both, amounts wagered and transactions are important. In 2008, the turnover generated by English bookmakers was about 10 billion pounds. Opening markets in Europe should see this trend increasing exponentially. Figure 1 shows the monthly variation in trades’ number of the biggest betting company, Betfair, compared on this statistic with financial markets. Betfair generates fewer trades than the CME but more than the LSE or NBOT. Its activity is comparable to an average-sized market as shown by the study of Roman (2007)1. Figure 1. Number of trades executed each month (in millions). The sports betting market may look distant from the financial markets. However, Barraud (2012) shows that it provides an experimental framework close enough in terms of organization, kind of participants and rules, to test the theory of informational efficiency. The existence of bets on sporting events creates opportunities for frauds, similar to those that can be found in the financial markets. It is very tempting to influence an athlete (or a team) to lose intentionally, and thus to collect gains without having taken risk. In doing so, the bettor has inside information: he knows the result before the other bettors and can make profits. The aim of this paper is to use existing similarities between financial markets and sports betting market to compare the detection of insider trading on financial markets and frauds in the sports betting market through the study of spread’s behavior. In doing so, the notions of strong form efficiency of financial markets on the one hand and fairness in sports betting market on the other hand will be discussed and compared. 2 Techniques development to identify insider trading is particularly interesting for both practitioners and theorists of financial markets. For practitioners, and particularly supervisors, they allow them to spot insider trading and to repress them, if they are strictly prohibited. Using the spread, organizers of sports betting may also easily identify suspicious games. For theorists, the identification of those operations allows to ensure on the one hand the reality of the strong form efficiency and on the other hand whether the use of privileged information is beneficial or not to investors. Nevertheless, testing such kind of market efficiency is tricky for searchers: when insider trades are prohibited, they are not all identified, except those suppressed by the supervisory authorities. When tests are made on the basis of these crimes, they are inevitably biased, we can easily think that the reactions of returns and trading volumes for insider trading which have not been identified by supervisors are different from those that have been identified, it is also perhaps why the first operations were identified and the others did not. This explanation is consistent with Kyle’s theoretical model (1985). This article is split into three parts. In the first part, we will present the notion of spread and specifications of sports betting. The second part will focus on the methodology and description of our sample. The results and key findings will form the third part. The spread Tests to detect insider trading are divided into three main groups: tests on abnormal returns, tests on abnormal volumes and tests on the spread. Tests on abnormal returns are both the best known and numerous. They are based on the well-known methodology of event study initiated by Fama, Fisher, Jensen and Roll. In their work, Meulbroeck (1992) and Guivarc'h (1997) use this methodology and do not validate the hypothesis of strong efficiency form, showing that using privileged information allow insiders to obtain abnormal returns. Insider trading can also be detected by observing, with a methodology quite similar, the existence of abnormal trading volumes. Fishe and Robe (2004) analyze transactions made by informed agents dealing on the Nyse, Nasdaq and Amex, and observe that transaction volumes increase on the day the insiders execute their orders. Finally, a more original detection of insider trading is based on the observation of the spread. The underlying assumption is while an insider is trading, spread may behave abnormally. Chung and Charoenwong (1998) show that in the case of a price driven market, market-makers widen the bid-ask spreads of stocks handled by the informed traders. Similarly, in a order driven market, Guivarc'h (1996) shows 3 that during an insider trading, short limited donors are placing orders in a less aggressive way, leading to a wider bid-ask spread. Two types of spread exist: the bid-ask spread also called the quoted spread and the effective spread. The quoted spread is defined as the difference between asking and biding price and is usually calculated in relative terms (% of assets), to allow comparison between stocks of which the price levels differs, as follows: Askt Bidt fat (1) Askt Bidt 2 The bid-ask spread assumes that transactions are always carried out on the best present limits in the order book, however, several authors including Huang and Stoll (1996) show that the best limits shown in the book do not correspond systematically to the price of transaction. This is the case on the Nasdaq, where some transactions are within the range when it is wide. In addition to the quoted spread, the effective spread takes into account the actual price of a transaction and is calculated as follows: Askt Bidt Pt 2 fet 2* (2) Askt Bidt 2 In this article we will therefore study the spread’s behavior as it may appear in the world of sports betting, we will seek to determine whether the conduct of the spread identified in insider trading on financial markets is reflected in the match-fixing in the sports betting market or not. Thus, if the spread’s behavior is the same in both worlds and in the same scenarios, the use of the spread to detect insider trading on financial markets or fraud in the sports betting market should be strengthened. The Betting Exchange The betting exchange is a type of bets which allows the bettor to play not against a bookmaker, but directly with other players. Bettors have the option to bet against the 4 execution of an event and to position themselves as sellers, in this case, they are called Layers. Thus, if we take the example of a horse race with five runners and we want to bet on the defeat of one of them, it is not necessary to bet on the winning of the other four, as would be the case in a conventional system, now, it’s possible to bet on the defeat of the horse who seems less able to win. This system simply allows any punter to take the place of the bookmaker. In this system, the odds will not be permanently fixed and will evolve according to changing market conditions, ie by financial flows corresponding to the players’ expectations. Like traditional financial markets, bettors can place multiple order types. Betfair Our sample is issue from Betfair, which is the largest online betting exchange in the world and the only one providing total access to data. Generally in the world of sports betting and more particularly on Betfair, competition mechanisms are close to those of financial markets and even surpasses them in terms of perfection: Betfair is the biggest company of sports betting, its popularity is the guarantee of a large number of punters, large volumes and thus liquidity. Information is free and is simultaneously revealed to all punters, for example in case of tennis matches, people have to watch TV to see if a player surpasses another or not. Everyone is able to see all orders, buying or selling. There is no hidden order2. Bet can be placed at any time before the start of an event or while the event unfolds, we call this period the Live Betting.